Machine Learning Engineer (Robotics)

Location: Hybrid in Munich or Berlin, Germany

Company

hirify.global is a Central European e-grocery company building an autonomous grocery operation across fulfilment, logistics, quality control, and physical robotics.

What you will do

• Define camera, mounting, calibration, synchronization, and episode-quality specifications for human-motion data capture.
• Own hand-pose ground truth and address occlusion, work gloves, cold warehouse conditions, and other real-world data challenges.
• Build the evaluation harness that determines whether captured hours are suitable for robot-policy training.
• Post-train open robot-learning models on the collected data and run initial task evaluations.
• Read and reproduce published methods, validating their performance before further data-capture investment.
• Work as the primary machine-learning specialist alongside an operations lead and data engineer.

Requirements

• Strong PyTorch and computer-vision experience with real 3D geometry.
• Experience with camera calibration, pose estimation, and SLAM fundamentals, including debugging extrinsic calibration.
• Ability to read research papers and reproduce published methods.
• Understanding of robot-learning policies and the data requirements for training them.
• Experience training models on self-collected datasets and diagnosing dataset failure modes.
• Experience with hand-pose estimation, egocentric video, multi-camera rigs, time synchronization, robot-learning model post-training, or ROS 2 is particularly relevant.

Culture & Benefits

• Join a small, newly created team focused on physical AI for autonomous grocery fulfilment.
• Own the capture specification and first training runs from the beginning.
• Build a specialised data asset from real grocery-warehouse operations.
• Use off-the-shelf hardware and a managed cloud environment.
• Direct technical judgment toward data-quality and capital-allocation decisions.
• Use AI coding agents such as Claude Code and Devin for implementation and data-engineering work while maintaining a high quality bar.

Hiring process

• Instead of a cover letter, explain where a selected human-video-to-robot-policy pipeline would fail in a warehouse where workers wear gloves.